[Bug]: Android work-log list crashes with "Maximum update depth exceeded" during active subagent work
メンテナーはふだん 1 日以内に返信
まだ誰も着手していません。
評価
- 難易度
- 4/5
- 見積もり時間
- 3〜5日
- 初心者へのやさしさ
- 52/100
- issue の種類
- バグ
- 明瞭さ
- おおむね明確
- 活発さ
- 活発
- 技術スタック
- react-native, typescript
調査の方向性
Start in the mobile app’s ThreadWorkGroupList and trace its use of AnimatedLegendList; the issue localizes the update-depth error there but does not identify a root cause. Compare the list’s subscription and layout updates while work-log activity streams, then reproduce on Android with an active helper if possible. Done means the thread view stays usable during streamed tool activity without the update-depth error.
索引モデルが issue の本文から書いたものです。
説明
Before submitting
- I searched existing issues and did not find a duplicate.
- I included enough detail to reproduce or investigate the problem.
Area
apps/mobile
Steps to reproduce
A deterministic minimal reproduction has not yet been isolated. Observed scenario:
- Use the T3 Code Android app to view a long-running thread connected to a Windows host.
- The thread has a running Claude Agent turn with an active helper/subagent producing tool activity.
- The mobile thread view fails with
Maximum update depth exceeded. The component stack runs throughThreadWorkGroupListand its nestedAnimatedLegendList, under aDisclosureRow.
The thread had approximately 2,633 top-level timeline items across 50 runs when inspected. The active helper had performed over 100 tool uses. These are context, not established reproduction requirements. The exact gesture or update immediately preceding the error (opening, expanding, scrolling, or a streamed update) is unknown.
Expected behavior
The Android thread view remains usable while the agent and its helper stream tool activity.
Actual behavior
React throws an update-depth error in the mobile work-log list.
Read-only inspection of the host immediately afterward confirmed that the affected run and its helper were still running, with fresh tool results and a subsequent test-game launch in the provider log. The failure therefore affected the client display while backend work continued.
No host restart, thread interruption, data deletion, or configuration changes were performed during diagnosis.
Impact
Major degradation or frequent failure
The affected Android thread view fails; backend work continues. Reproduction frequency has not been established.
Version or commit
Windows host desktop/server installation: 0.1.9-nightly.20261110.4465, read from the host's desktop-last-started-version.
Android client version/build and commit: not yet captured. The host version should not be treated as the Android client version.
Environment
- Client: native T3 Code Android app (
com.t3tools.t3code), Expo bundle. - Phone model and Android version: not yet captured.
- Host: Windows; T3 Code Nightly.
- Provider: Claude Agent, with an active helper/subagent.
- Observed November 10, 2026, around 19:15–19:17 UTC.
Logs or stack traces
Relevant excerpt of the user-visible error below. Identical repeated frames are abbreviated; bundle offsets are retained.
Error: Maximum update depth exceeded. This can happen when a component repeatedly calls setState inside componentWillUpdate or componentDidUpdate. React limits the number of nested updates to prevent infinite loops.
at enqueueConcurrentRenderForLane (bundle:1:2656200)
at updateStoreInstance (bundle:1:2662076)
at commitHookEffectListMount (bundle:1:2694735)
at commitPassiveMountOnFiber (bundle:1:2707897)
at recursivelyTraversePassiveMountEffects (bundle:1:2706866)
... repeated passive-mount traversal frames; original trace skips 602 frames ...
at flushPassiveEffects (bundle:1:2727716)
at flushPendingEffects (bundle:1:2727499)
at flushSpawnedWork (bundle:1:2727214)
at commitRoot (bundle:1:2726179)
at completeRootWhenReady (bundle:1:2713817)
at performWorkOnRoot (bundle:1:2713082)
at flushSyncWorkAcrossRoots_impl (bundle:1:2652976)
at anonymous (bundle:1:7828190)
Component stack:
at ContainersLayer2 (bundle:1:6621412)
at Containers2 (bundle:1:6621928)
at RCTView
at View (bundle:1:2598473)
at RCTScrollView
at ScrollView (bundle:1:2946037)
at ScrollView (bundle:1:2949105)
at ScrollView (bundle:1:2911254)
at AnimatedComponent(ScrollView) (bundle:1:7956644)
at ScrollView (bundle:1:7956701)
at AnimatedScrollView (bundle:1:3146147)
at ReanimatedScrollBridgeComponent (bundle:1:6555513)
at ListComponent2 (bundle:1:6622933)
at LegendListInner2 (bundle:1:6626035)
at StateProvider (bundle:1:6559227)
at LegendList2 (bundle:1:6625770)
at LegendListForwardedRef2 (bundle:1:6557277)
at AnimatedComponent(Component) (bundle:1:7956644)
at Component (bundle:1:7956701)
at AnimatedLegendList2 (bundle:1:6557680)
... View wrappers ...
at ThreadWorkGroupList (bundle:1:7017735)
... View wrappers ...
at WorkLogBlock (bundle:1:7042214)
at ThreadWorkLog (bundle:1:7015012)
at ThreadMediaVisibility (bundle:1:7189355)
at MessageImageGalleryScope (bundle:1:7204101)
... View / animated View wrappers ...
at DisclosureRow (bundle:1:7202939)
... View wrappers ...
at PositionViewState2 (bundle:1:6618328)
at Container2 (bundle:1:6619875)
at ContainerSlotBase (bundle:1:6605679)
at ContainerSlot2
Here bundle stands for:
/data/user/0/com.t3tools.t3code/files/.expo-internal/42bac505bc3eda455134bc0b498d30dc.
Investigation notes
The stack localizes the failure to the grouped work-log list, but does not establish the root cause. A list snapshot/subscription or layout feedback loop is a hypothesis, not a confirmed diagnosis. No source patch or Android reproduction was attempted.
Related reports found during duplicate search:
- #21866: Android update-depth crash in the sidebar/inspector path; this trace instead contains
ThreadWorkGroupList. - #15704: mobile grouped tool-call scrolling problem in the same component, but a different symptom.
Workaround
No workaround has been verified on the affected Android client. Using desktop/web or collapsing the work-log group was suggested, but not tested. The running host and thread were deliberately left untouched.
Private conversation text, thread/environment identifiers, local project paths, and raw provider logs are omitted.
- 主要言語
- TypeScript
- スター
- 24.8k
- フォーク
- 6.4k
- 平均マージ
- 10時間 56分
- マージ済み PR(30日)
- 226
環境構築
このプロジェクトの開発コンテナを、あなたの GitHub アカウントでブラウザ上に起動します。
- Dockerfile・Docker Compose ファイルなし
- プルリクエストのテンプレートあり
- コントリビューションガイドを読む
はじめの一歩
- issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
- 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
- リポジトリをフォークし、ブランチを切って変更します。
- issue 番号を参照したプルリクエストを送ります。
pingdotgg/t3code のほかの issue
-
[Bug]: request_secret file delivery has no trailing newline, and the tool text does not say so対応中かも このイシューにリンクされたプルリクエストがオープン中、またはマージ済みです。 オープン
難易度 2/5 1〜3時間 初心者へのやさしさ 69/100
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 68/100
メンテナーはふだん 1 日以内に返信
-
[Bug]: Changing the sidebar project filter collapses an expanded Settled shelf back to 10 threadsオープン
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
pingdotgg/t3code#34291 · コメント 2 件 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 83/100
メンテナーはふだん 1 日以内に返信
-
via-triage
難易度 2/5 1〜3時間 初心者へのやさしさ 76/100
メンテナーはふだん 1 日以内に返信
pingdotgg/t3code の issue をすべて見る
似ている issue
-
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
メンテナーはふだん 1 日以内に返信
-
community first-timers-only good first issue hacktoberfest help wanted low hanging fruit up-for-grabs
難易度 1/5 1時間未満 初心者へのやさしさ 82/100
lingdojo/kana-dojo#33947 · コメント 1 件 · リアクション 5 件 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
paperclipai/paperclip#20942 ·
メンテナーはふだん 1 日以内に返信
-
area: documentation good first issue package: rule-data status: accepting prs
難易度 2/5 1〜3時間 初心者へのやさしさ 72/100
メンテナーはふだん 1 日以内に返信
-
status: needs triage
難易度 2/5 1〜3時間 初心者へのやさしさ 85/100
mastra-ai/mastra#30109 · コメント 2 件 ·
メンテナーはふだん 1 日以内に返信